runner.runner-disk-prune-ec2-vm

Run docker-prune and workspace-temp cleanup on a warm-pool EC2 VM runner via SSM. Fire-and-forget hook called from execution-stop to keep disk-pressure flat between jobs (ADR-018).

Run docker-prune + workspace temp cleanup on a single EC2 VM.

Triggered by runner.execution-stop-ec2-vm once the GH job is confirmed finished. The workflow itself does no synchronisation — SSM SendCommand is asynchronous and we don't wait for the in-VM exit, so the workflow completes in a single round-trip per instance. If SSM denies the call (missing IAM, instance offline, not SSM-managed) we log + complete success: the reconcile loop's disk-pressure rule remains the safety net.

Inputs:

• runner_instance_uuid — DB row identifying the runner to prune. Resolved server-side to get the cloud connection + region.

Outputs (state_data):

• runner_instance_uuid — echoed back
• ssm_command_ref — opaque SSM CommandId on success, None on skip / failure
• status — issued | skipped | failed

Overview

Property	Value
Workflow type	Linear
Library	App-runners-aws
Version	1.0

Input Schema

Field	Type	Required	Default	Description
runner_instance_uuid	uuid	Yes	—	UUID of the RunnerInstance to prune, picked from the organization's (optionally group-scoped) list of instances. Resolved server-side to get the cloud connection + region.

Output Schema

Field	Type	Required	Default	Description
runner_instance_uuid	uuid	No	—	UUID of the RunnerInstance that was pruned, echoed from input.
ssm_command_ref	string	No	—	—
status	string	No	—	—
completed_at	string	No	—	—
failed_at	string	No	—	—
failure_reason	string	No	—	—
error	string	No	—	—
error_type	string	No	—	—
failed_at_state	string	No	—	—
failed_layer	string	No	—	—
failed_step	string	No	—	—

States

State	Initial	Terminal	Success	Auto-advance	Description
pending	Yes	No	—	prune	—
pruning	No	No	—	complete	—
completed	No	Yes	Yes	—	—
failed	No	Yes	No	—	—

State Diagram

Transitions

From	Action	To	Description
pending	prune	pruning	—
pruning	complete	completed	—
* (any state)	fail	failed	—

API Usage

POST /api/workflows/start
Content-Type: application/json

{
  "workflow_type": "runner.runner-disk-prune-ec2-vm",
  "initial_data": {
    "runner_instance_uuid": "value"
  }
}